I personally prefer OpenGL over Direct3d. One of the biggest benefits of OpenGL is that it is open sourced where DirectX is commercial. Both are consided required for serious graphic programmer but it seems to me that OpenGL is losing favor with developers. In the past(DirectX 5 era) OpenGL was considered more powerful that DirectX however it seems that alot of developers think that this isn't the case. I have read alot of postings that state that OpenGL will lose it's current support from graphics and game developers when DirectX 10 comes. I personnally refuse to think that OpenGL will lose over Direct even if there are more books and more people that support it OpenGL will always be open source.

opengl is a specification, i think

AFAIK, opengl isnt "open source". however, i believe there are open source implementations of it, but the one im thinking of (mesa?) is a software renderer.

however, i generally found it easy to learn opengl, where as the insane amount of setting up and handles that need to be learned put my off directx.

the major benefit of opengl is that it is cross platform, asar as im concerned.

the other issue is directx is a graphics/network/input/otherstuff api, while opengl is a grahpics api.

they aren't directly comparable.

but SDL/openGL i find easier on the head. my 2 cents

In any well designed engine, toggling between D3D and OGL (at least right now) is a trivial proposition. Sure it might require some QA, but the basic port is quick.

Now, with D3D 10, OpenGL/ARB will have to tack on some kludges...excuse me, approve some extensions quickly in order to come up to snuff, or they will have serious problems. OpenGL is no longer driving innovation, as it was for several years. They're playing a pure catch-up game now, and if they mess it up, they'll be relegated to the sidelines (i.e. alternate OSes).

All things considered, OpenGL already is "relegated to the sidelines." There's no real compelling reason to use OGL if all your developing for is Windows, and since that's 99%-ish of the PC game market very few developers are going to do ports anyways. For someone in it purely for commerical reasons, DirectX is definately more attractive.

That said, like Promit mentioned: What API an engine runs on is pretty insignificant anymore. A well designed engine should be able to port from one to the other in a negligable amount of time. The days are gone where an engine would be pitched as "DirectX based" or "OpenGL based". It's all about the feature set and the tools now.

OpenGL is never going to "die": there are too many modeling apps and the like that use it, and too many older games that need it. The card manufacturers will continue to support it, regaurdless of how badly MS neglects it, and it will continue to be used for ports and various consoles. By and large, though, DirectX is the standard now, and that's the way it looks to stay for a long time.

(Note: I develop primarily using OpenGL, so don't take me to be a DirectX fanboy or anything. I'm just stating the facts.)

Ah, the "OpenGL is going to diieee!!!11" post of the week. It was already overdue.

*sigh*

Anyway. Will OpenGL lose the "battle" against DX10 ? Nope, it won't. Let's look at some facts, shall we ? First, an API doesn't "drive innovation" anymore, these times are long over. The hardware vendors drive innovation, no, they push it at all price. Both major APIs just try to keep up with it.

And MS has an impressive track record of completely failing at this "catch the new GPU feature" game. The DX9 instancing fiasco is a perfect example where hardware vendors started rebelling against MS's API policies, by even technically circumventing it. MS learned their lesson (well, partially at least), and made DX10 broader in feature context than what has been done before.

OpenGL doesn't work this way, because it uses a different release model. OpenGL evolves continously, while D3D does it in jumps. It has always been the same: DX is a step ahead of OpenGL when a new version is released. Then after some time both are on par again. And finally, OpenGL gets the lead because new GPU features get released as GL extensions. Repeat the cycle for the next version.

So yeah, DX10 has features current OpenGL implementations don't have. So what ? OpenGL will get them as soon as the demand is there. And just wait until GPUs have surpassed DX10 features. Then OGL will lead again, until DX11. And so on. It's just the way both update models work.

Quote:
There's no real compelling reason to use OGL if all your developing for is Windows,

And there is no real compelling reason not to use it.

Quote:
and since that's 99%-ish of the PC game market very few developers are going to do ports anyways. For someone in it purely for commerical reasons, DirectX is definately more attractive.

How so ?

In fact, it's more a question about the graphics framework game companies license from third parties. Most companies won't give a shit about whether the engine they buy is D3D or OGL, as long as it's cutting edge and not too expensive. The reason why currently more 3D engines in the game sector use D3D is historical (well, around 2, 3 years ago). Theoretically, this could change again at any time.

Some misunderstanding people got from my original post, and it's entirely my fault for just putting in another attention getting & poorly written OpenGL post. Here is one post by the mod Yann L that shows how I screwed up my argument for OpenGL losing popularity.

Quote:
Ah, teh "OpenGL is going to diieee!!!11" post of the week. It was already overdue.

*sigh*

Well I never did say that OpenGL was going to "diieeee!!11" what I was trying to get at is how alot of developers are using Direct3d where openGL would work better. Take Carmack for instance he was a long term OpenGL guy and on the latest Doom installment he worked with DirectX 8 stating that DX8 at the time had the best pixel shader support and the best vertex shader support.

Perhaps I am just overeacting to Microsoft's success with Direct. But I wish that more individuals would avoid DirectX just because Microsoft is a 10 ton corperation with a "mostly" bad rep.

Also on a side note I also f**ked up by stating that OpenGL is open source, Opengl by itself isn't, I should have said it is the current industry standard. I was actually thinking about mesa which is an open source implementation of the OpenGL. I also forgot to state that opengl is cross-platform and that too many computer users are trapped in to using windows when they have other options. BTW sorry about my piss poor writtening style and all my english errors.

Quote:Original post by Hodge
Some misunderstanding people got from my original post, and it's entirely my fault for just putting in another attention getting & poorly written OpenGL post. Here is one post by the mod Yann L that shows how I screwed up my argument for OpenGL losing popularity.

I wasn't specifically referring to your post. This is yet another standard D3D versus OGL thread. Whatever the exact way of starting it might be, it will always end in the same way, trust me on that. People always use the same fallacious arguments, ignore facts, and twist reality to fit with whatever API they might prefer. Again, this is not targeted at you in particular, it's just a typical property of such threads: objectivity dies first.

Quote:
Well I never did say that OpenGL was going to "diieeee!!11" what I was trying to get at is how alot of developers are using Direct3d where openGL would work better. Take Carmack for instance he was a long term OpenGL guy and on the latest Doom installment he worked with DirectX 8 stating that DX8 at the time had the best pixel shader support and the best vertex shader support.

Doom 3 uses OpenGL, except for the XBox version (obviously).

Quote:
Perhaps I am just overeacting to Microsoft's success with Direct. But I wish that more individuals would avoid DirectX just because Microsoft is a 10 ton corperation with a "mostly" bad rep.

And that is supposed to be a valid reason to avoid it ? Select an API based on objective technical assessments. Select it based on your personal preference for its semantics. Select it based on driver, OS or platform support. But avoiding an API because it is done by "the big evil M$" is just stupid. No offense intended, but I smell zealotry from miles around.

Quote:
I also forgot to state that opengl is cross-platform and that too many computer users are trapped in to using windows when they have other options.

Windows is the industry standard on consumer level platforms right now. Deal with it. As an indie developer, you have the choice: use Windows or Linux, use D3D or OGL. Use what you prefer, or what might be better suited to your particular needs. Choice is always good.

But if you are a large scale game development studio, your priority is not to bring down Windows or D3D and introduce the next OS revolution. Your priority is to get the game done on schedule, to reduce costs, and to make shareholders and/or publishers happy. What API is used to do that is completely irrelevant. More often than not, the techies (who might be able to judge the pros and contras of an API) are not even involved in the selection process. The purchasing and finance department will choose whatever engine seems to be most suited in terms of minimizing investment and a fast ROI.

In the industry, there are no OGL versus D3D arguments. The question doesn't even exist. It's all about the bottom line. Create a groundbreaking new OpenGL based 3D engine that sells cheaper than existing engines, and suddendly everybody will use OpenGL. Do the same with D3D, and the balance tips into the other direction again. Right now, it's the latter. Tomorrow, everything can change again.